Unobtrusive Sensing Technologies for Supporting Remote Monitoring in the Home

Abstract

My thesis focuses on exploration and implementation of various sensing technologies that support in-home activity recognition. The technologies presented in this thesis include (1) a wearable device and a platform for studying the elders acceptance of carrying an attachment, (2) an IMS-based technique for capturing fine-grained characteristics of electrical events that can be attributed to human activities and (3) a technique to enable unmodified LCD monitors to sense human proximity and hand gestures. In particular, I have shown that human behaviors can cause time-varying EMI (electromagnetic interference), which can be sensed from a single set of sensing hardware installed anywhere in the home. These granular characteristics open a new feature space to explore insights of electrical events and can be used to enable various applications such as activity inference, energy disaggregation and motor failure detection. In addition, the IMS-based approach presented in this dissertation can potentially support continuous remote monitoring, acting as an automatic diary system, for helping the elderly increase their awareness of daily activities and in the meantime provides the rehabilitation researchers a low-cost method to track the patients activeness while the patients are in the home.